I use the different following functions to compute model coverage: cvtest, cvsim, decisioninfo. In my models I use blocks of a library that has been developped by my company and that I have to use.
The problem is that these blocks of library are made with switch to have particular options: for instance a specific "Divide" block has been created with an option that allows the user to switch the 2 inputs (x and /). This means that there is a switch that is hidden behind the "Divide" block and when I compute model coverage, even with my inputs, I am going through this "Divide" block I have only 50% in decision.
Is there a way to prevent model coverage from looking into masked subsystems? Or to prevent model coverage from looking into some blocks?
Thank you for your help.